ChangeLog

-----------------------------------------------------------------------------------------
$Id: ChangeLog 679 2017-06-15 03:11:54Z pwessel $
	Contains changes made to the data files under GSHHS or WDBII
	Note: As of late 2012 the subversion project is called gshhg. It used to
	be called GSHHS+WDBII.  Make sure you use gshhg.
	
-----------------------------------------------------------------------------------------
2017-06-15	GSHHG 2.3.7 released
-----------------------------------------------------------------------------------------
2017-06-10	Replace all of Northern Mariana Islands with CUSP data from NOAA after
		simplifying down to ~50 meter resolution to match GSHHS full level
			
2017-06-08	Added two missing islands in Norway: Sula and Ellingsøya with the
		new data from Statens Kartverk via Birgitte R. Furevik [birgitterf@met.no]
			
2017-04-25	Added Kosovo border with Serbia
-----------------------------------------------------------------------------------------
2015-08-19	GSHHG 2.3.6 released
-----------------------------------------------------------------------------------------
2016-08-06	pwessel
	Revised data for Jan Mayen, Norway, thanks to Norwegian Polar Institute, via
	https://data.npolar.no/dataset/63730e2e-b7a6-4d14-b341-c661ccdc5254.
2016-08-06	pwessel
	Added missing islands Georgetown and MacMahan, ME, thanks to Duncan Agnew, Scripps
2016-06-18	pwessel
	Fixed 11 crossings in Antarctica grounding line and one in icefront, thanks to
	David Redhouse, University of Cambridge
-----------------------------------------------------------------------------------------
2015-04-12	GSHHG 2.3.5 released
-----------------------------------------------------------------------------------------
2016-02-22	pwessel
	Fixed non-closure of the Slovenia, Croatia and Hungary border.

2015-04-09	pwessel
	Added missing boundary between Sudan and South Sudan.
-----------------------------------------------------------------------------------------
2015-01-01	GSHHG 2.3.4 released
-----------------------------------------------------------------------------------------
2014-12-29	pwessel
	Added missing Swedish Lake Maelaren and flipped 11 "lakes" to islands instead.
-----------------------------------------------------------------------------------------
2014-11-01	GSHHG 2.3.3 released
-----------------------------------------------------------------------------------------
2014-10-29	pwessel
	Removed the Saudi-Kuwait neutral zone which went out of style in 1991...
-----------------------------------------------------------------------------------------
2014-08-01	GSHHG 2.3.2 released
-----------------------------------------------------------------------------------------
2014-07-15	pwessel
	The crossover checker skipped the two new Antarctica polygons so there were a few tens of
	crossings.  These have now been removed.
-----------------------------------------------------------------------------------------
2014-02-01	GSHHG 2.3.1 released
-----------------------------------------------------------------------------------------
2014-06-29	pwessel
	During the switch to the new Antarctica data we lost Peter I Island.  It has now been updated
	and included in the database.
-----------------------------------------------------------------------------------------
2014-04-10	pwessel
	Due to a bug in polygon_to_gshhg.c the Antarctica polygons had an unswapped n
	value which became junk in the binary file.  I fixed the bug and rebuilt the
	binary files only.  No other changes so no version changes.
-----------------------------------------------------------------------------------------
2014-02-01	GSHHG 2.3.0 released
-----------------------------------------------------------------------------------------
	New gshhg-gmt-2.3.0* archives built.  Only the 5.2 branch of GMT takes advantage
	of the choice for Antarctica coastlines; earlier versions can only see the ice-
	front boundary.  However, the gshhg supplements of all GMT versions will output
	the polygons, using new levels 5 and 6 for ice- and grounding-line polygons.
-----------------------------------------------------------------------------------------
2013-11-07	pwessel
	Removed all GSHHG Antarctic islands from Level_1 files and added in new islands
	from Bohlander and Scambos (2007). There are now TWO Antarctic coastlines; one ice
	line and the grounding line.  Lines and islands are given source = 2 for ice-front
	and source = 3 for grounding line (and associated islands).  These are given the new
	levels 5 and 6, respectively, in order to be backwards compatible with older code.
	While gmt_shore.c was updated it has some bugs so we need more debugging and testing
	before we can release 2.3.0.  So GMT 5.1.0/4.5.11 was released for use with 2.2.4.
	These bugs have now been fixed, so 5.2.x will be compatible with 2.3.0 while 5.1.x
	and 4.5.x are backwards compatible with 2.3.0 in the sense that the grounding line
	coastline is invisible to these older versions.
	
-----------------------------------------------------------------------------------------
2013-11-01      GSHHG 2.2.4 released
-----------------------------------------------------------------------------------------
2013-10-22      pwessel
        One lake in Antarctica was actually a bad copy of an island.  Removed.

2013-08-06      pwessel
        Found that a few key lakes in California (Lake Mono (+ two islands), Trinity,
        Isabella) were missing, so these were added. Also found a bug in polygon_consistency
        that failed to fine some spikes (~20-25 polygons affected).

-----------------------------------------------------------------------------------------
2013-07-01	GSHHG 2.2.3 released
-----------------------------------------------------------------------------------------
2013-06-01	pwessel
	With input from Tom Kratzke we found ~120 spikes with width < 2 meters that were
	removed.  This affected a few dozen full resolution polygons.  Also fixed an error
	in Baffin Island in all but the crude resolutions; this modification also converted
	two mislabeled "lakes" into islands in the full and high resolution files. We then
	removed almost 200 skinny polygons (3-4 point polygons with area < 0.1% of the
	full resolution size equivalent) and fixed a few spikes in the h, i, l, c resolutions.

2013-01-06	pwessel
	Because GSHHG is now released under the lesser GNU license we have added
	those two COPYING* files and a modified LICENSE.TXT, similar to GMT.
	The new files will appear in the next distribution.

-----------------------------------------------------------------------------------------
2013-01-01	GSHHG 2.2.2 released
-----------------------------------------------------------------------------------------

2012-12-27	pwessel
	Based on the Agalega case I checked for WDBII islands and found 112.  Of these,
	48 were imprecise versions of more accurate WVS features and were removed.  These
	were mostly found in the Red Sea, the Persian Gulf, and in the Cook-Austral region.
	There are remaining inaccuracies associated with the Ganges Delta, but I don't have
	access to better data.

2012-12-23	pwessel
	Alert user Philippe Veerabadren, Mauritius pointed out that the N and S Agalega
	Islands appears twice in GSHHG.  Turned out both the WDBII and WVS versions were
	in there.  I removed the WDBII versions and shifted the WVS by (+0.01435, -0.011283)
	to better fit the Google Earth satellite data.

2012-11-23	pwessel
	Following Seton et al Nov 2012 discovery that Sandy Island, Coral Sea, is not real
	(well, not anywhere near or above sea level according to multiple satellites) we
	have removed Sandy island from full (Id=1203), high (1203), intermediate (1203)
	and low (1201) resolutions (it was not present in crude due to its small size).
	We now release GSHHG under the lesser GNU Public License.  This allows users
	to employ GSHHG in their code without having to release their source back.

2012-10-10	pwessel
	Made minor adjustments to Tahiti and Moorea plus their tiny reefs and islets (we
	subtracted 0.0155 from longitude) and replaced Mehetia, courtesy of Keith Pickering.
	Made new lower-res versions from these full-resolution polygons.

-----------------------------------------------------------------------------------------
2012-05-01	GSHHG 2.2.1 released
-----------------------------------------------------------------------------------------
	We have renamed the product to GSHHG - Global Self-consistent Hierarchical
	High-resolution Geography, reflecting the fact the the distribution contains
	both shoreline polygons AND borders/rivers. This affects the naming of the
	distributed packages; only minor changes to the data has occurred since 2.2.0;
	see comments below.

2012-01-31	pwessel
	Fixed minor glitch in SA-NT border.

2011-07-30	Under subversion control.
	Deleted 7 zero-length line segments from the 3 border files.  They were sets
	of 2 identical points each.  Automatically skipped when building the binary
	files but now also deleted from the archive.

	-----------------------------------------------------------------------------------------
2011-07-15	GSHHS 2.2.0 released
-----------------------------------------------------------------------------------------

2011-06-27	pwessel
	Store polygon area using adjustable magnitude scale so that smaller features
	can retain precision (and features smaller than 0.1 km^2 no longer being seen
	as having no area ).  Furthermore, the areas were recomputed using the WGS-84
	ellipsoid as the previous values were based on a spherical calculation.  Thanks
	to José Luis García Pallero for bringing this issues to our attention.
	Also adjusted the greenwich flag to be 1 if crossing Greenwich (as before) but 2
	if crossing Dateline, and 3 if crossing both (only Eurasia and Antarctica);
	otherwise set to 0.  Custom code can now use greenwich to adjust longitudes.
	Finally, the binary gshhs files now store Antarctica in -180/+180 range so as
	to avoid a longitude jump when dumped to ASCII.  Also, the WDBII shapefiles
	only had the first 3 levels of rivers; version 2.2.0 has all 11.  Finally,
	to be able to detect the river-lake features in the WDBII binary files we
	set the river flag to 1 if a closed feature.

-----------------------------------------------------------------------------------------
2011-03-10	GSHHS 2.1.1 released
-----------------------------------------------------------------------------------------

2011-01-08	pwessel
	Added missing segment of border between Germany and Poland ending at the
	Baltic Sea. Removing 4 spikes in full and one spike in the intermediate
	coastline files [Thanks again to Tom Kratzke].
	
2010-12-20	pwessel
	Fixing a few minor problem in the low resolution file related to the manual
	editing of 2010-07-16.  Plus a few more non-area spikes as well. This includes
	pols 6367, 6453, and 6469 in the low resolution file.  This will result in a
	revised GSHHS version 2.1.1.

2010-07-16	pwessel
	To aid in the dumb splitting of polygons crossing the Dateline (for GIS stuff),
	I manually added in the crossing coordinates for lon == 180.  This applied to
	5-6 polygons per resolution.  polygon_to_shape.c can now do a very simple splitting.

-----------------------------------------------------------------------------------------
2010-07-15	GSHHS 2.1.0 released
-----------------------------------------------------------------------------------------

2010-06-16	pwessel
	With feedback (and data) from Branislav Mikulas (Branislav.Mikulas@iblsoft.com) I
	have added some missing borders to the WDBII borders, level 1: Syria-Israel,
	Israel-Jordan, Moldova-Ukraine, and the Eritrea border with Ethiopia.  Seems OK.

2010-05-16	pwessel
	(a) After restoring the negative areas for river-lakes I made the binary files, then
	ran polygon_sort on these to restore the proper sorting, then converted back to
	ascii with binary_to_ascii, then reformatted the binary files with "make data".
	(b) Running polygon_consistency revealed several polygons in the i, l, c resolutions
	with zero-angle excursions so I removed those polygons and reran step (a) above.
	(c) Since polygon IDs have changed due to the removal of features and resorting
	I ran polygon_findlevel for all five resolutions, then polygon_hierarchy to find
	the parent IDs, and polygon_sync to update the metadata.  Then, again did binary_to_ascii
	and updated the ASCII files which were then updated in cvs.
	
2010-04-30	pwessel
	Sometime during or after GMT 4.5.2 I made the mistake of removing the sign of the
	area for river-lakes.  As these were marked by having a negative area we lost the
	ability to identify them.  The fix was to extract the most recent cvs version that
	still had the negative areas, get those header records only, and use awk to write a
	script that would apply sed to change "A = " to "A = -" for those records only.
	This was done only for the five res_?/GSHHS_?_Level_2.txt files since it is a lake
	issue only.
	
2009-06-05	pwessel
	Reverted the river database using GMT 4.3.1 version to restore whatever
	was screwed up during the August-28-1007 activity listed below.
	Corrected a dozen minor issues pointed out by Tom Kratzke, including an
	island (level 3) that was listed as lake in a lake.  Corrected gross errors
	in the coastline near the mouth of the Amazon, resulting in numerous lakes
	becoming islands and vice versa.  Extended two riverlakes to connect to the
	coastline.  Other riverlakes flagged with negative area so they can be given
	special treatment by programs.  Now store the parent id of polygon containing
	a polygon (-1 for level 1 features).  At this point, only the full resolution
	has these corrections and improvements.

2007-08-28	pwessel
	Note on rivers: The original WDBII classification for rivers was this:
	01      Permanent major rivers
	02      Additional major rivers
	03      Additional rivers
	04      Minor rivers
	05      Double lined rivers
	06      Intermittent rivers -- major
	07      Intermittent rivers -- additional
	08      Intermittent rivers -- minor
	10      Major canals
	11      Canals of lesser importance
	12      Canals -- irrigation type

	However:
	1. 05 double rivers was manually stitched into long skinny lakes and added
	   to the coastline database back in mid/late 1990ies.  Since this classification
	   was not always well done in the original data we now add the double rivers
	   to class 1 (permanent major rivers); this yields better plots.
	2. 12 Irrigation canals:  The files seems to skip 12 and use 13 for this type.

	Thus, when GMT prepares the cdf files we change the level values as follows:
	1-4 pass through
	5 becomes 1 as well
	6-8 becomes 5-7
	10-11 becomes 8-9
	13 becomes 10

2007-06-03	pwessel
	Several polygons have zero area in the original GMT polygons.  We
	need to recalculate these.

2007-06-01	pwessel
	Need to verify that the binary files produced from ascii via
	ascii_to_polygon do match the latest GMT polygons.

	Change how area is handled: Use irint(log10(area-in-km)*2e8)
	instead to preserve precision.  This would mean changes to
	polygon_to_bins.c, shoremaker.c, and gmt_shore.c.  Perhaps add
	a new area netcdf variable so that old code can run on new files.
	Bring WDBII source lines (borders and rivers) into CVS as well.
	The full rivers are 21 Mb binary - split into ???